400B
Parameters
MoE (17B active per expert)
AIVory Smart Inference · Llama
Open WeightLlama 4 Maverick - 400B quality, startup pricing.
Meta's flagship open-weight model. 400B parameters, 17B active per expert, native image input. One base_url swap and you're running it.
At a glance
Llama 4 Maverick via Smart Inference.
Routed to the cheapest live provider on every request. Same model, lower bill.
128K
Context window
tokens per request
—
Cheapest input price
per 1M tokens via Smart Inference
—%
Cheaper
vs the priciest endpoint for the same model
Pricing
Llama 4 Maverick: price comparison.
Our cheapest routed price vs what you'd pay elsewhere.
Prices in USD per 1M input tokens · updated live from the router
Quick start
Use Llama 4 Maverick in three lines.
OpenAI-compatible. Swap your base_url, keep everything else.
from openai import OpenAI
client = OpenAI(
base_url="https://smart.aivory.net/v1",
api_key="sk-aivory-...",
)
resp = client.chat.completions.create(
model="llama-4-maverick",
messages=[{"role": "user", "content": "Ship it."}],
)
print(resp.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://smart.aivory.net/v1",
apiKey: process.env.AIVORY_API_KEY,
});
const resp = await client.chat.completions.create({
model: "llama-4-maverick",
messages: [{ role: "user", content: "Ship it." }],
});
console.log(resp.choices[0].message.content);curl https://smart.aivory.net/v1/chat/completions \
-H "Authorization: Bearer $AIVORY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "llama-4-maverick",
"messages": [{"role": "user", "content": "Ship it."}]
}'That's the whole migration. Get early access and try it →
Use cases
What Llama 4 Maverick is great at.
-
Multimodal Document Processing
Send a 50-page PDF screenshot alongside a prompt. Maverick reads the image, extracts tables, and returns structured JSON. The 131K context window fits the full document plus your system prompt.
-
Multi-Step Analysis
Feed a codebase diff and ask for security review, test suggestions, and a summary in one call. With 17B active parameters per expert, Maverick routes internally to the right specialists for each subtask.
-
Long-Form Generation
Draft a 5,000-word technical report with citations from a 80K-token source document. The MoE architecture keeps output coherent across long runs without the cost of a full 400B dense forward pass.
FAQ
Common questions.
What does Llama 4 Maverick cost through AIVory?
The price depends on which provider the router picks. As of June 2026, the cheapest input rate is around $0.13 per million tokens. The live table above shows the current rate. No subscription, no minimum spend.
Can Maverick process images?
Yes. Maverick accepts image inputs natively. Pass base64-encoded images in the messages array, same format as the OpenAI vision API. The router only selects providers that support multimodal input for image requests.
Maverick vs Llama 4 Scout - when to use which?
Scout (109B, 17B active) costs less and responds faster. Maverick (400B, 17B active per expert, more experts) handles harder reasoning and image tasks. Start with Scout. Move to Maverick when Scout's output quality isn't enough.
How many providers host Maverick?
Multiple cloud providers including Azure, AWS, GCP, and RunPod serve Maverick endpoints. The router evaluates all of them on every request and picks the cheapest healthy one. You don't manage provider accounts.
What license is Llama 4 Maverick released under?
Meta's Llama 4 Community License. Open-weight with commercial use allowed. You can also self-host it on a spot GPU through the GPU Marketplace if you need dedicated capacity.
Start using Llama 4 Maverick. Today.
Swap your base_url, keep shipping.
The router is live. Credits from $10, no subscription, no commitment.